Development of Biodiesel Production Processes from Various Vegetable Oils

Development of Biodiesel Production Processes

from Various Vegetable Oils

A thesis

submitted to the college of graduate studies and research in partial fulfillment of the requirements for the degree of

Doctor of Philosophy (Ph.D.)

in the Division of Environmental Engineering University of Saskatchewan

Saskatoon, Saskatchewan

by

Titipong Issariyakul

i

Permission to Use

In presenting this thesis in partial fulfillment of the requirements for a doctorate of philosophy degree from the University of Saskatchwan, I agree that the libraries of the University of Saskatchewan may take this thesis freely available for inspection. I also agree that permission for extensive copying of this thesis for scholarly purposes may be granted by Dr. Ajay K. Dalai, who supervised this thesis work recorded herein, or in his absence, by the Graduate Chair of the Division of Environmental Engineering or the Dean of the College of Graduate Studies. It is understood that due recognition will be given to the author of this thesis and to the University of Saskatchewan in any use of the material of the thesis. Any copying, publication, or use of the thesis or part of this thesis for financial gain is prohibited without my written permission.

Requests for permission to copy or to make other use of material in this thesis in whole or in part should be addressed to:

Graduate Chair

Division of Environmental Engineering University of Saskatchewan

Saskatoon, Saskatchewan Canada S7N 5A9

ii

Acknowledgement

This thesis would have not been possible without guidance and assistance of several individuals who one way or another appreciably contributed to the completion of this thesis.

First of all, I would like to express my earnest and sincere gratitude to my supervisor, Dr. Ajay K. Dalai not only for his valuable guidance, encouragement, and support but also for giving me an opportunity to train and produce research work during the course of my Ph.D. program. I am also would like to convey my deep appreciation to my co-supervisor Dr. Narendra N. Bakhshi for his continuous support and thoughtfulness.

I cordially thank my advisory committee members, Dr. J. Cutler, Dr. M. Nemati, Dr. R. Ranganathan, and Dr. M. Reaney, for their valuable comments which helped to improve the quality of this thesis. I appreciate technical assistance and precious discussions from visiting professors and my colleagues at the Catalysis and Chemical Reaction Engineering Laboratories as well as technicians and secretaries including R. Blondin, H. Eunike, J. Compain, J. Horosko, K. Bader, D. Cekic, in the Division of Environmental Engineering and the Department of Chemical and Biological Engineering, University of Saskatchewan.

My deepest gratitude goes to my family especially my mom Mrs. Duangporn and my aunt Ms. Na Keaw for their everlasting love and support that helped me to overcome obstacles during the course of my Ph.D. study. Last but not the least; I warmly thank Kung for her continuous love and friendliness that always rejuvenates my spirit.

iii Abstract

Biodiesel is an alternative fuel to petroleum diesel that is renewable and creates less harmful emissions than conventional diesel thus the use of this fuel is a shift toward “sustainable energy”. Biodiesel can be produced from vegetable oil, animal fat, and organisms such as algae or cyanobacteria. Since vegetable oils are the major source for current commercial biodiesel, they are the focus of this thesis.

The main objective of this Ph.D. research is to develop processes suitable to produce biodiesel from various vegetable oils especially for those of non-edible oils such as used cooking oil, canola oil from greenseed, and mustard oil. An additional objective is to understand the relationship between the parent vegetable oils and the corresponding biodiesel properties.

Used cooking oil was the first vegetable oil investigated in this research. Initially, oil degradation behavior was monitored closely during frying. During 72 hours of frying, acid value and viscosity of the oil increased from 0.2 to 1.5 mgKOH·g-1 and from 38.2 to 50.6 cP, respectively. It was found that ester yield was improved by addition of canola oil to used cooking oil, i.e. addition of 20% canola oil to used cooking oil increased methyl ester yield and ethyl ester yield by 0.5% and 12.2%, respectively. At least 60% canola oil addition is needed to produce ASTM grade ethyl ester biodiesel. The optimum reaction conditions to produce biodiesel are 1% KOH loading, 6:1 alcohol to oil ratio, 600 rpm stirring speed, and either 50°C reaction temperature for 2 hr or 60°C reaction temperature for 1.5 hr for methanolysis and 60°C reaction temperature for 2 hr for ethanolysis.

Among non-edible vegetable oils, greenseed canola oil can be used in the most simple biodiesel production process. In this case, an addition of fresh vegetable oil is not required,

iv

because chlorophyll contained in this oil did not play a crucial role in the reaction activity. Methyl ester yields derived from greenseed canola oil without and with 94.1 ppm chlorophyll content are 95.7% and 94.8%, respectively. In contrast, erucic acid contained in mustard oil created difficulties in the production process. Ester yield derived from mustard oil using the conditions mentioned above was only 66% due to the present of unconverted monoglyceride. To obtain a deeper understanding on mustard oil transesterification, its reaction kinetics was studied. In the kinetic study, transesterification kinetics of palm oil was also investigated to study the effect of fatty acid chain length and degree of saturation on the rates of the reactions. It is shown in this research that the rates of mustard monoglyceride transesterification (rate constant = 0.2-0.6 L·mol-1·min-1) were slower that those of palm monoglyceride transesterification (rate constant = 1.2-4.2 L·mol-1·min-1) due to its lower molecular polarity resulting from the longer chain of erucic acid. The activation energy of the rate determining step (in this case, conversion of triglyceride to diglyceride reaction step) of mustard transesterification was, however, 26.8 kJ·mol-1, which is similar to those of other vegetable oils as reported in literature. Despite the presence of unconverted monoglyceride, distillation can be used to obtain a high purity ester.

Several ester properties are determined by characteristics of the parent oil and choice of alcohol used in transesterification. Chlorophyll contained in greenseed canola oil, for example, has an adverse effect on biodiesel oxidative stability. The induction time for methyl ester derived from treated greenseed canola oil (pigment content = 1 ppm) was enhanced by 12 minutes compared to that derived from crude greenseed canola oil (pigment content = 34 ppm). The optimum bleaching process involves the use of 7.5 wt.% montmorillonite K10 at 60°C and stirring speed of 600 rpm for 30 minutes. In addition, it was found that induction time of treated greenseed canola ethyl ester (1.8 hr) was higher than that of methyl ester (0.7 hr), which suggests

v

a better oxidative stability of esters of higher alcohols. Furthermore, the use of higher alcohols instead of methanol produced materials with improved low temperature properties. For example, the crystallization temperatures of monounsaturated methyl, ethyl, propyl, and butyl esters prepared from mustard oil were -42.5°C, -51.0°C, -51.9°C, and -58.2°C, respectively. In contrast, the lubricity of biodiesel is mainly provided by its functional group which is COOCH3 for methyl ester. The use of higher alcohols in transesterification results in a less polar functional group in the corresponding ester molecule, which leads to reduction in ester lubricity. Methyl ester provided the highest lubricity among all esters produced, i.e. wear reduction at 1% treat rate of methyl ester, ethyl ester, propyl ester, and butyl ester are 43.7%, 23.2%, 30.7% and 30.2%, respectively.

The outcomes of this research have been published in several scientific journals and presented at national and international conferences. The published articles and conference presentations are listed at the beginning of each chapter in this thesis.

vi Table of Contents Permission to Use i Acknowledgement ii Abstract iii Table of Contents vi

List of Tables xiv

Lest of Figures xvii

CHAPTER 1

Introduction and Research Overview 1

1.1. Introduction 1.2. Research Overview References 1 2 3 CHAPTER 2 Literature Review

Contribution of the Ph.D. Candidate

Contribution of this Chapter to the Overall Ph.D. Research

4 4 4 2.1. Abstract 2.2. Introduction 2.3. Feedstock 2.3.1. Soybean Oil 5 5 7 12

vii 2.3.2. Rapeseed Oil, Mustard Oil, and Canola Oil 2.3.3. Palm Oil

2.3.4. Sunflower Oil 2.3.5. Rice Bran Oil 2.3.6. Jatropa Oil 2.3.7. Karanja Oil 2.3.8. Used Cooking Oil 2.4. Biodiesel Production

2.4.1. Effects of Free Fatty Acid and Water Content 2.4.2. Effects of Alcohol

2.4.3. Effects of Catalyst Type

2.4.3.1. Homogeneous Base Catalysis 2.4.3.2. Homogeneous Acid Catalysis 2.4.3.3. Heterogeneous Base Catalysis 2.4.3.4. Heterogeneous Acid Catalysis

2.4.4. Effects of Reaction Time, Temperature, and the Reaction Kinetics 2.4.5. Techniques for Monitoring Transesterification

2.4.5.1. Gas Chromatography 2.4.5.2. Liquid Chromatography

2.4.5.3. Nuclear Magnetic Resonance Spectroscopy 2.4.5.4. Infrared Spectrometry

2.4.5.5. Other Methods 2.4.6. Post Reaction Treatment

18 20 21 21 22 22 23 27 27 31 34 35 42 44 50 56 59 60 64 67 68 69 70

viii 2.5. Biodiesel Quality 2.5.1. Combustion Properties 2.5.2. Flow Properties 2.5.3. Stability 2.5.4. Lubricity 2.5.5. Minor Components 2.5.5.1. Pigments

2.5.5.2. Lecithin and Phospholipids 2.5.5.3. Phytosterols

2.5.5.4. Glycolipids

2.6. Biodiesel Production in Canada 2.7. Conclusions References 71 80 81 83 86 89 93 94 95 96 96 100 101 CHAPTER 3

Oil Degradation during Frying and its Effects on

the Corresponding Biodiesel Yield and Oxidative Stability Contribution of the Ph.D. Candidate

Contribution of this Chapter to the Overall Ph.D. Research

126 126 126 3.1. Abstract 3.2. Introduction 3.3. Materials 3.4. Experimental Procedures 127 127 129 129

ix 3.4.1. Frying Process for Extended-Life Canola Oil 3.4.2. Transesterification

3.4.3. Characterization 3.5. Results and Discussions

3.5.1. Oil Degradation during Frying

3.5.2. Transesterification and Ester Analysis 3.6. Conclusions Abbreviations References 129 130 131 131 131 134 138 140 141 CHAPTER 4

Biodiesel Production from Canola Oil and Used Cooking Oil Contribution of the Ph.D. Candidate

Contribution of this Chapter to the Overall Ph.D. Research

143 144 144 4.1. Abstract 4.2. Introduction 4.3. Materials 4.4. Experimental Procedures 4.4.1. Transesterification 4.4.2. Characterization 4.5. Results and Discussions

4.5.1. Feedstock Analysis 4.5.2. Product Analysis 145 145 147 147 147 149 150 150 152

x 4.5.3. Characterizations of the Esters

4.5.4. Ester Purification

4.5.5. Thermal Analysis by Differential Scanning Calorimetry (DSC) 4.6. Conclusions Abbreviations References 158 159 161 168 170 172 CHAPTER 5

Biodiesel Production from Greenseed Canola Oil Contribution of the Ph.D. Candidate

Contribution of this Chapter to the Overall Ph.D. Research

174 174 175 5.1. Abstract 5.2. Introduction 5.3. Materials 5.4. Experimental Procedures

5.4.1. Bleaching of Greenseed Canola Oil 5.4.2. Transesterification

5.4.3. Characterization 5.5. Results and Discussions

5.5.1. Bleaching 5.5.2. Transesterification 5.5.3. Biodiesel Properties 5.5.4. Oxidative Stability 176 176 178 179 179 180 182 183 183 187 190 194

xi 5.6. Conclusions Abbreviations References 198 200 202 CHAPTER 6

Biodiesel Production from Mustard Oil Contribution of the Ph.D. Candidate

Contribution of this Chapter to the Overall Ph.D. Research

205 205 205 6.1. Abstract 6.2. Introduction 6.3. Materials 6.4. Experimental Procedures 6.4.1. Transesterification

6.4.2. Distillation of Transesterification Products 6.4.3. Characterization

6.5. Results and Discussions 6.5.1. Feedstock Analysis

6.5.2. Transesterification Products of Mustard Oil

6.5.3. Mixtures of Esters derived from Canola, Soybean and Mustard Oil 6.5.4. Esters Production from Mustard Oil

6.5.5. Distillation of Transesterification Products Derived from Mustard Oil 6.5.6. Characterization of Mustard Biodiesel

6.5.7. Lubricity 206 206 209 210 210 213 213 215 215 216 217 218 223 224 229

xii 6.5.8. Oxidative Stability 6.5.9. Crystallization 6.6. Conclusions Abbreviations References 231 232 234 236 238 CHAPTER 7

Transesterification Kinetics of Vegetable Oils Contribution of the Ph.D. Candidate

Contribution of this Chapter to the Overall Ph.D. Research

240 240 241 7.1. Abstract 7.2. Introduction 7.3. Materials 7.4. Experimental Procedures 7.4.1. Feedstock Analysis 7.4.2. Transesterification 7.4.3. Sample Preparation 7.4.4. Characterization 7.4.5. The Kinetic Model 7.4.6. The MATLAB Program 7.5. Results and Discussions

7.5.1. The Mass Transfer Effect 7.5.2. Repeatability 242 242 246 247 247 249 249 251 252 254 256 256 258

xiii 7.5.3. The Rate Constants and Activation Energies 7.6. Conclusions Abbreviations References 259 267 269 270 CHAPTER 8

Conclusions and Recommendations 273

8.1. Conclusions 8.2. Recommendations 273 276 APPENDIX A HPLC Calibration 277 APPENDIX B

FAME Standard for GC Analysis 283

APPENDIX C

GC Calibration 284

APPENDIX D

MATLAB Program for Transesterification Kinetics 287

APPENDIX E

xiv List of Tables

Table 2.1 World oilseed production, average oil price and oil content

of various oilseeds 9

Table 2.2 Molecular structure of triglyceride, diglyceride, and monoglyceride 10 Table 2.3 Structures of common fatty acids found in vegetable oils 10 Table 2.4 Fatty acid compositions of vegetable oils 13 Table 2.5 Iodine value and saponification value of vegetable oils 17 Table 2.6 Examples of homogeneous catalysis on esterification and transesterification 38 Table 2.7 Examples of heterogeneous catalysis on esterification and transesterification 47 Table 2.8 Fuel standards and test methods for pure biodiesel 72 Table 2.9 Fuel standards ASTM D7467 for B6 to B20 and CAN/CGSB-3.520

for B1 to B5 blended biodiesel-petroleum diesel fuel 77 Table 2.10 Biodiesel plants in Canada 97 Table 3.1 Induction time and acid value of the feedstock oil 134 Table 3.2 Fatty acid compositions of RBD canola oil methyl ester (CME), greenseed

canola oil methyl ester (GME), and used cooking oil methyl ester (UME) 135 Table 3.3 Product compositions of esters produced from RBD canola, greenseed

canola, and used cooking oil (transesterification at 60°C for 1.5 h) 136 Table 4.1 Triglyceride, diglyceride, and monoglyceride percentage in esters 156 Table 4.2 Characteristics of esters 157 Table 4.3 Boiling point distribution (°C) of esters 160 Table 4.4 Characteristics of purified ethyl esters 161 Table 4.5 Fatty acid compositions of canola oil methyl ester

xv

Table 4.6 Major peak temperature and heat associated to crystallization

and melting of esters 167

Table 5.1 Biodiesel samples 181

Table 5.2 Initial pigment content and acid value in canola oil and greenseed canola oil 183 Table 5.3 Physical properties and performance of various adsorbents for

pigment adsorption 184

Table 5.4 Properties and performance of regenerated montmorillonite K10 185 Table 5.5 Performance of montmorillonite K10 at various bleaching durations

for pigment adsorption 186

Table 5.6 Performance of montmorillonite K10 at various percent loading

for pigment adsorption 187

Table 5.7 Percent (w/w) of triglyceride, diglyceride, monoglyceride, ester,

and percent ester recovery 189

Table 5.8 Fatty acid compositions of selected esters 191 Table 5.9 Acid value, iodine value, viscosity @40°C and sulfur content of esters 193 Table 5.10 Pigment content of esters 197 Table 6.1 Fatty acid compositions and AV of S. alba and B. juncea 211 Table 6.2 Mixtures of esters produced from canola, soybean and mustard oil 212 Table 6.3 Analysis of transesterification products 214 Table 6.4 AV, water content, ester and glyceride content of vegetable oils 216 Table 6.5 Water content, AV, ester and glyceride content of biodiesel

derived from canola, soybean, and mustard oil 219 Table 6.6 Fatty acid compositions of mustard, canola, and soybean oil methyl ester 223 Table 6.7 Properties of mustard oil and biodiesel produced from mustard oil 226

xvi

Table 7.1 Fatty acid composition of palm oil and mustard oil 248 Table 7.2 Chemical properties of palm oil and mustard oil 249 Table 7.3 The rate constants of each reaction step during transesterification 264 Table 7.4 Activation energy of the rate determining step of transesterification 267 Table E1 Prices of feedstocks for biodiesel production 296

xvii List of Figures

Figure 2.1 Scheme for oxidative reaction mechanism 24 Figure 2.2 Carbon-hydrogen bond positions in fatty acids 25 Figure 2.3 Scheme for step-wise transesterification reaction 29 Figure 2.4 Hydrolysis and saponification during transesterification:

a) saponification of free fatty acid; b) saponification of triacylglyceride;

c) hydrolysis of methyl ester 30 Figure 2.5 Mechanism for homogeneous base catalysis in transesterification 36 Figure 2.6 Mechanism for homogeneous acid catalysis in esterification

and transesterification 43

Figure 2.7 Mechanism for heterogeneous base catalysis in transesterification 45 Figure 2.8 Mechanism for heterogeneous acid catalysis in esterification

and transesterification 51

Figure 2.9 Chromatographic separation of component A and B

and their corresponding output signals 61 Figure 2.10 Chemical solution model for biodiesel blends 89 Figure 2.11 Minor components in vegetable oils: a) chlorophyll; b) α-tocopherol

or vitamin E; c) β-carotene; d) phospholipid; e) sterol; f) sterol glycoside (β–sitosterol-β-D-glucopyranoside);

g) glycolipid; h) glucosinolate 92 Figure 2.12 Chlorophylls degradation pathways 94 Figure 3.1 Acid value of extended-life canola oil during the frying process 132 Figure 3.2 Viscosity @40°C of extended-life canola oil during the frying process 133 Figure 3.3 Change in ester percentage during transesterification of used cooking oil,

greenseed canola oil and RBD canola oil 136 Figure 3.4 Oxidative stability plot of RBD canola oil methyl ester (CME) 137

xviii

Figure 3.5 Induction time of esters produced from used cooking oil,

greenseed canola oil, and RBD canola oil 138 Figure 4.1 HPLC chromatograms: a) canola oil; b) used cooking oil;

c) canola oil methyl ester; d) used cooking oil methyl ester 151 Figure 4.2 Ester percentage as analyzed by HPLC analysis: ● methyl ester;

▲ ethyl ester (reaction conditions: alcohol to oil ratio 6:1,

temperature 50°C) 153

Figure 4.3 Amounts of ester and glycerol collected from transesterification of 100 g of feedstock: a) ester recovery; b) glycerol recovery; ● methanolysis; ▲ ethanolysis (reaction conditions: alcohol to oil ratio 6:1,

temperature 50°C) 154

Figure 4.4 Typical DSC thermogram of canola oil methyl ester 163 Figure 4.5 Cooling curves of DSC thermogram: a) used cooking oil methyl ester;

b) 40:60 UCO:CO methyl ester; c) canola oil methyl ester 166 Figure 5.1 Ester formation during transesterification

of canola oil using methanol at 50 and 60°C 188 Figure 5.2 A conductivity-time plot for oxidative stability determination of canola oil

methyl ester 195 Figure 5.3 Induction time of esters prepared from canola oil and greenseed canola oil 196 Figure 5.4 Ester formation during transesterification of greenseed canola oil

and pigment content of ester at the end of the reaction 198 Figure 6.1 Transesterification reaction scheme 207 Figure 6.2 Ester and glyceride content during transesterification of mustard oil 217 Figure 6.3 Ester formation during transesterification of mustard oil using CH3ONa

and KOH as a catalyst 220

Figure 6.4 Ester formation during transesterification of mustard oil at elevated

temperature and pressure (150°C, 3.4 MPa) 221 Figure 6.5 Boiling point distribution of canola and mustard oil methyl ester using

xix

Figure 6.6 HPLC chromatograms of (a) undistilled mustard oil methyl ester; (b) distilled mustard oil methyl ester; and (c) residual from

distillation of mutard oil butyl ester 225 Figure 6.7 Lubricity properties of methyl esters derived from mustard oil and canola oil

blended with reference diesel fuel using

high frequency reciprocating rig (HFRR) method 230 Figure 6.8 A conductivity-time plot for oxidative stability determination of esters

derived from mustard oil 231 Figure 6.9 DSC thermogram of esters derived from mustard oil 233 Figure 7.1 Scheme for step-wise transesterification reaction 244 Figure 7.2 Scheme for transesterification based on shunt reaction mechanism 246 Figure 7.3 Effect of stirring speed during palm oil transesterification at 60°C on

a) TG conversion using 200 and 600 rpm;

b) TG conversion at 1 minute using 200, 400, 600, and 800 rpm 257 Figure 7.4 Concentrations change during palm oil transesterification at 60°C

and 600 rpm: a) triglyceride concentration; b) diglyceride concentration;

c) monoglyceride concentration; d) methyl ester concentration 260 Figure 7.5 Concentrations of the reaction mixture during

a) palm oil; b) mustard oil transesterification at 60°C and 600 rpm 261 Figure 7.6 Experimental and simulated data during palm oil transesterification at 60°C:

a) triglyceride concentration; b) diglyceride concentration; c) monoglyceride concentration; d) methyl ester concentration;

e) glycerol concentration; f) methanol concentration 262 Figure 7.7 Experimental and simulated methyl ester concentrations during

a) palm oil; b) mustard oil transesterification 263 Figure 7.8 Arrhenius plots of the rate determining step:

a) palm oil transesterification; b) mustard oil transesterification 266 Figure A1 Chromatogram of a triolein standard 277 Figure A2 Chromatogram of a diolein standard 278 Figure A3 Chromatogram of a monoolein standard 278

xx

Figure A4 Chromatogram of a methyl oleate standard 279 Figure A5 Chromatogram of a standard mixture 279 Figure A6 Chromatogram of a standard mixture with glycerol 280 Figure A7 Triglyceride calibration curve 280 Figure A8 Diglyceride calibration curve 281 Figure A9 Monoglyceride calibration curve 281 Figure A10 Ester calibration curve 282 Figure B1 Chromatogram of FAME standard (10 mg/mL FAME in methylene chloride) 283 Figure C1 Triglyceride calibration curve 284 Figure C2 Diglyceride calibration curve 285 Figure C3 Monoglyceride calibration curve 285 Figure C4 Ester calibration curve 286

1

CHAPTER 1

Introduction and Research Overview

1.1. Introduction

Biodiesel is an alternative renewable diesel fuel that has properties comparable to diesel obtained from petroleum processing. Since biodiesel is renewable and it creates less harmful exhaust emissions when combusted compared to that of petroleum diesel, the use of this fuel is a shift towards “sustainable energy”. Biodiesel can be produced from vegetable oil, animal fat, or organisms such as algae and cyanobacteria through a chemical reaction called transesterification with short chain alcohols. Since vegetable oils are currently the major source of feedstock in commercial biodiesel production, the focus of this Ph.D. research is biodiesel production based on vegetable oils. In addition to alternative fuel, biodiesel is commonly viewed as a lubricity additive to petroleum diesel. Because biodiesel is miscible with petroleum diesel in all proportions, an addition of only 1 vol.% biodiesel to petroleum diesel improves the lubricating property of petroleum diesel [1]. Due to its various advantages especially environmental benefits, many governments worldwide encourage the use of this fuel in the form of tax incentives and mandate implementations.

In 2011, the Canadian government implemented a 2% federal mandate for biodiesel that creates demand for around 500 million litres per year of biodiesel in Canada. According to Canadian Renewable Fuels Association, the domestic biodiesel production capacity around Canada was 200 million litres per year in 2010 [2]. There is an obvious demand for domestic

2

biodiesel production boost in order to reduce dependency on imported biodiesel. Therefore, major growth in biodiesel industry is expected in the coming years and research and technology on biodiesel production processes will be of higher value than ever. Research done in this thesis will serve as information for researchers and biodiesel manufacturers both inside and outside Canada. More details on feedstock, production processes, and characteristics of biodiesel are extensively reviewed in Chapter 2.

1.2. Research Overview

The objective of this research is to investigate biodiesel production from various vegetable oils with special emphasis on non-edible oil. In addition, the effects of different oils used as feedstock on reaction activity as well as the resulting biodiesel properties are also studied. This research is divided into 5 phases which are discussed in Chapter 3 to Chapter 7 and are outlined as follow.

In Chapter 3, properties of used cooking oil are monitored closely during frying and biodiesel is produced from used cooking oil. The properties of biodiesel derived from used cooking oil are compared with those of biodiesel prepared from canola oil and greenseed canola oil. In an attempt to improve properties of used cooking oil biodiesel, used cooking oil is mixed with canola oil and the mixed oil is used as feedstock for biodiesel production. This process is described in Chapter 4. The optimum mixing ratio is reported along with the optimum reaction conditions. Furthermore, Differential Scanning Calorimetry (DSC) is used to measure crystallization and melting temperature as well as heat associated with crystallization and melting. In this study, the effects of lipid and alcohol feedstock used in transesterification on crystallization and melting behaviour of biodiesel are reported.

3

Chapter 5 is focused on biodiesel production from greenseed canola oil. The effects of chlorophyll and its derivatives on transesterification activity as well as biodiesel property are discussed with special emphasis on oxidation stability. The optimum conditions on bleaching and transesterification are reported. Biodiesel production from mustard oil is then investigated in Chapter 6. A production process to produce high quality mustard biodiesel is developed. Biodiesel lubricity is evaluated and compared to those of commercial petroleum diesel. Finally, transesterification kinetics is investigated. Mathematical model as well as MATLAB program are developed in order to simulate transesterification progress and kinetic parameters such as the rate constants and the activation energies are obtained. Transesterification kinetics of different vegetable oils are compared and discussed in Chapter 7.

Since this thesis is provided in the paper-base format, each chapter has been published in national and international scientific journals and this is mentioned at the beginning of each chapter and the contribution of the Ph.D. candidate is highlighted. Abbreviations and references are given at the end of each chapter.

References

[1] Lang X, Dalai AK, Reaney MJ, Hertz PB. Biodiesel esters as lubricity additives: effects of process variables and evaluation of low-temperature properties. Fuels International: 207-227 (2001).

[2] Canadian Renewable Fuels Association. Growing beyond oil delivering our energy future: a report card on the Canadian renewable fuels industry. www.greenfuels.org. Accessed November 2010.

4

CHAPTER 2

Literature Review

A part of this chapter has been submitted for publication in Renewable & Sustainable Energy Reviews:

• Issariyakul, T., and Dalai A.K. Biodiesel from vegetable oils. Renewable and Sustainable Energy Reviews (Submitted).

Contribution of the Ph.D. Candidate

Literature review was performed by Titipong Issariyakul. The content in this chapter was written by Titipong Issariyakul with discussions and suggestions provided by Dr. Ajay Dalai.

Contribution of this Chapter to the Overall Ph.D. Research

In Chapter 2, history of the use of vegetable oil related fuel is reviewed along with feedstock, production, and characteristics of biodiesel from vegetable oils. This chapter provides the overall direction of this Ph.D. research.

5 2.1. Abstract

Biodiesel is gaining acceptance in the market as both fuel and lubricant. It is expected that the biodiesel industry will grow rapidly worldwide in the coming years and information regarding biodiesel feedstock, its production, and characteristics will be significant. Biodiesel from vegetable oil is the focus in the present review. Since vegetable oil is currently the major source for making commercial biodiesel, selected available vegetable oils are reviewed as feedstock for biodiesel production. Production technologies including the latest catalyst developments are discussed and biodiesel characteristics and parameters influencing the corresponding biodiesel properties are revealed.

2.2. Introduction

As conventional, non-renewable, fossil-based fuel resources are depleting, research and development on alternative renewable energy is growing. Biodiesel is a promising renewable energy. Recently, a 5.54 fossil energy ratio (FER) is reported [1] which means one unit of fossil energy input is required to produce 5.54 units of biodiesel energy output from soybean oil. This FER indicates a superior energy return of biodiesel that surpasses those of other alternate fuels [2]. The FER of biodiesel is expected to increase in the coming years, due to increased crop yield, adoption of energy-saving farm practices, and continuous development technologies that enhance energy efficiency. Biodiesel has many properties that contribute to superior performance when compared to petroleum diesel. For example, biodiesel produces lower exhaust emissions than conventional diesel and it is biodegradable, non-toxic, renewable, and essentially free of

6

sulfur [3,4,5]. Since biodiesel is renewable and environmentally beneficial, the use of this fuel is a shift towards sustainable energy.

The history of biodiesel is as long as that of the diesel engine itself and the use of vegetable oils was investigated as early as the era when diesel engine was developed. Rudolf Diesel (1858-1913), the inventor of diesel engine, tested peanut oil as fuel for his engine. Dating from early 1920s, many vegetable oils were investigated, including palm oil, soybean oil, cottonseed oil, and castor oil. These early studies showed satisfactory performance of vegetable oil as fuel for diesel engines [6]. However, there were concerns that their higher costs as compared to petroleum fuel would prevent their prevalent uses. In spite of their performance in diesel engine, vegetable oils create engine problems when used as diesel fuel in both indirect- and direct-injection engines. The major drawback of vegetable oils is their high viscosity which causes coking, varnishing and trumpet formation on the injectors that results in poor atomization and ultimately leads to operational problems such as engine deposits [7].

Possible solutions to reduce the viscosity of vegetable oil include heating, transesterification, pyrolysis, dilution with petroleum-based fuels, and emulsification [8]. Transesterification is the most common method which yields mono alkyl esters of long chain fatty acids or fatty acid alkyl ester (FAAE). This idea originated in 1938 when it was noted that glycerine has a low calorific value and is likely to cause excess carbon deposit in the engine and, therefore, should be eliminated from glyceride oils used as diesel fuel. During that period, it was proposed that the engine should run on what was referred to as “residue fatty acid” [9]. This residue fatty acid is known today as “biodiesel”, although ester was not yet mentioned. In fact, the high molecular weight of the triglyceride molecule is responsible for much of the viscosity of vegetable oil, whereas the fatty acids are typically 10 times less viscous than their parent

7

vegetable oil at room temperature. During the summer of 1938, an urban bus running between Brussels and Louvain was operated on ethyl ester produced from palm oil. The engine performance was satisfactory and it was noted that the viscosity of ethyl ester was less than that of palm oil. The term “biodiesel” made its first appearance in a paper published in 1988 and this term was used exponentially thereafter [10].

Biodiesel is often defined as the mono alkyl esters of long chain fatty acids. Such esters may be prepared from acyl-glycerides (usually triglyceride) in vegetable oils via transesterification with short chain alcohols. Biodiesel is miscible in all portions with petroleum based diesel and, thus, can be effectively used as a neat biodiesel or blended with petroleum based diesel fuel [11]. The blends of biodiesel and petrodiesel are often coded to represent the percent volume of the blend. B20, for example, indicates the blend of 20 vol.% biodiesel and 80 vol.% petrodiesel. The current knowledge of biodiesel feedstock chemistry (vegetable oils), transesterification reactions, and biodiesel properties are described in the following sections.

2.3. Feedstock

Both lipid and alcohol feedstock determine the properties and method used in biodiesel production. Lipid feedstocks include vegetable oils, animal fats, and, more recently, oil from other organisms such as micro algae and cyanobacteria [12,13]. This paper focuses on vegetable triglyceride oils as lipid feedstock. The vegetable oils available for biodiesel production highly depend on climate. Rapeseed oil is utilized in European countries and Canada, soybean oil in the United States of America, and palm oil in tropical countries including Indonesia and Malaysia while coconut oil is used for synthesis of biodiesel in coastal areas. Potential inedible oils used as

8

lipid feedstock include jatropha oil (Jatropha curcas) and karanja oil (Pongamia pinnata) [14]. Oilseed prices and availability are important parameters to consider as biodiesel feedstock and are shown in Table 2.1.

Soybean and palm oil dominate world oilseed production while considerably less rapeseed oil production occurs. The oil content of rapeseed is >40%, while that of soybean is 21%. Palm oil is an interesting source for biodiesel production due to its low price and relatively high oil content (40%). Oil palm also achieves a higher annual oil yield compared to soybean and rapeseed.

Oilseeds store lipid in organelles called oleosomes which can be broken and extracted to produce vegetable oil. The major component of vegetable oils is triacylglycerol (TAG) or triglyceride (TG) which is a molecule composed of three esters of fatty acid chain (acyl group) attached to glycerol (glycerol group). When two acyl ester groups and one hydroxyl group (– OH) are present, the molecule is called a diacylglycerol (DAG) or diglyceride (DG). Similarly, monoaclyglycerol (MAG) or monoglyceride (MG) has one acyl ester group and two hydroxyl groups. Acylglycerol is a term referred to TAG, DAG, or MAG and is depicted in Table 2.2. The acyl groups are typically unbranched fatty acids with between 10 to 24 carbon atoms. Saturated fatty acids have no double bonds between carbon atoms. When a pair of hydrogen atoms are removed from a fatty acid chain, one double bond is present and, therefore; it is called monounsaturated fatty acid. If the molecule contains two or more double bonds caused by further removal of hydrogen atoms, it is called polyunsaturated fatty acids. These fatty acids are frequently represented by a symbol such as C18:1, which indicates a fraction consisting of 18 carbon atoms and one double bond. Typical fatty acids attached to TAG found in vegetable oils are presented in Table 2.3.

Table 2.1 World oilseed production, average oil price and oil content of various oilseeds.

a_{Data in 2006/2007;} b_{Canola oil}

Plant Oil content

(%)

Oilseed productiona (Million metric tons)

Average oilseed pricea (U.S.D/metric ton)

Average oil pricea (U.S.D/metric ton) Yield (kg oil/hectare/yr) Reference Rapeseed >40 46.72 375 852b 600 - 1000 15,16,17 Soybean 21 235.77 254 684 300 - 450 15,16,18

Sunflower seed 44-51 30.15 n/a n/a 280 - 700 15,16,19

Palm 40 10.27 n/a 655 2500 - 4000 15,16,20

Cottonseed 18 46.02 n/a 787 n/a 15,20

Peanut 36-56 32.36 395 1253 340 - 440 15,16,21

Copra 65-68 5.28 537 n/a n/a 15,22

Coconut 63 n/a n/a 812 600 - 1500 15,16,20

10

Table 2.2 Molecular structure of triglyceride, diglyceride, and monoglyceride.

Triglyceride Diglyceride Monoglyceride

R1, R2, R3 = fatty acid chain

Table 2.3 Structures of common fatty acids present in vegetable oils. System name Common name Symbol Formula Double bond

positiona

Saturated

Decanoic Capric C10:0 C10H20O2 - Dodecanoic Lauric C12:0 C12H24O2 - Tetradecanoic Myristic C14:0 C14H28O2 - Hexadecanoic Palmitic C16:0 C16H32O2 - Octadecanoic Stearic C18:0 C18H36O2 - Eicosanoic Arachidic C20:0 C20H40O2 - Docosanoic Behenic C22:0 C22H44O2 - Tetracosanoic Lignoceric C24:0 C24H48O2 - Monounsaturated

Hexadecenoic Palmitoleic C16:1 C16H30O2 9c Octadecenoic Petroselinic C18:1 C18H34O2 6c

H2C   HC H2C OH O O C C R2 R3  O=  O=  H2C    HC  H2C  O  O  O  C  C  C  R1  R2  R3  O=   O=    O=    H2C   HC  H2C  OH  OH  O  C  R3  O=   

11

Octadecenoic Oleic C18:1 C18H34O2 9c Octadecenoic Asclepic C18:1 C18H34O2 11c Eicosenoic n/aa C20:1 C20H38O2 5c Eicosenoic Gadoleic C20:1 C20H38O2 9c Eicosenoic Gondoic C20:1 C20H38O2 11c Docosenoic Erucic C22:1 C22H42O2 13c

Polyunsaturated

Hexadecadienoic n/aa C16:2 C16H28O2

Octadecadienoic Linoleic C18:2 C18H32O2 9c12c Octadecatrienoic Linolenic-α C18:3 C18H30O2 9c12c15c Octadecatrienoic Linolenic-γ C18:3 C18H30O2 6c9c12c Octadecatrienoic Eleostearic C18:3 C18H30O2 9c11t13t Octadecatrienoic Calendic C18:3 C18H30O2 8t10t12c

a_{c = cis formation; t = trans formation; n/a = not available}

Stereo isomers of unsaturated fatty acids can be arranged in cis and trans orientation. Most natural occurring fatty acids from vegetable oils have cis-double bonds whereas the unnatural

trans-isomers usually only occur due to partial hydrogenation process. The Latin prefixed cis and trans describe the orientation of hydrogen atoms attached to carbon atoms at position next to a

double bond. In cis-isomer, hydrogen atoms are attached on the same side causing a “V” shape of fatty acid chain.

12

The major difference between various vegetable oils is the type of fatty acids attached in the triglyceride molecule. Fatty acid compositions of various vegetable oils are provided in table 2.4. Fatty acid composition determines biodiesel fuel properties, therefore, vegetable oil fatty acid composition is important [42]. Both the degree of saturation/unsaturation and molecular weight of vegetable oils determine fuel properties. The degree of saturation/unsaturation is proportional to the iodine value while the saponification value is inversely proportional to molecular weight. Iodine value and saponification value of selected vegetable oils are provided in Table 2.5 [43].

2.3.1. Soybean Oil

Glycine max is referred to as “Soybean” or “Soya”. This member of Papilionaceae is found only

under cultivation. The origin of soybean is not clear, for the genus Glycine has two major gene centres; eastern Africa and Australia. It is believed that the genus Glycine was dispersed from Australia to the whole Pacific region including China via migratory birds as seed carriers. Based on historical and geographical evidence, north eastern China is considered to be the site of soybean domestication. There are a number of soy-based food products including various liquids prepared from soybean and soybean curd known as “tofu”. From China, soybean spread through nearby countries including Korea, Japan, and the Southeast Asian region. More recently, soybean has been cultivated around the world. Soybean was first mentioned in USA literature in 1804. From that time until World War II, it was exclusively used as a forage crop, after which its production and economic value in the USA has grown exponentially. Today, soybean is the world’s largest oilseed in terms of total production and international trade [44,45].

Table 2.4 Fatty acid compositions of vegetable oils.

Vegetable oils Fatty acid composition (wt.%)

Reference Common Name Species 12:0 14:0 16:0 16:1 18:0 18:1 18:2 18:3 20:0 22:0 22:1

Canola (Low

erucic rapeseed) Brassica rapa - - 3.1 0.2 1.3 56.6 22.4 14.0 0.4 0.2 0.1 23

Canola (Low

erucic rapeseed) Brassica napus - - 4.3 0.3 1.7 61.0 20.8 9.3 0.6 0.3 - 23

Black mustard Brassica nigra

- 1.5 5.3 0.2 1.3 11.7 16.9 2.5 9.2 0.4 41.0 24

Oriental mustard Brassica

juncea

- - 2.3 0.2 1.0 8.9 16.0 11.8 0.8 5.7 43.3 25

Brown mustard Brassica juncea

- - 2.2 0.2 1.2 17.4 20.5 14.1 0.7 0.5 28.1 26

Wild mustard Sinapis arvensis

- 0.1 2.6 0.2 0.9 7.8 14.2 13.0 0.8 1.5 45.7 27

White mustard Sinapis alba - - 3.1 0.2 0.7 9.1 11.7 12.5 0.7 - 46.5 26

White mustard Sinapis alba - 0.1 2.8 0.2 1.1 25.0 11.6 8.6 0.7 0.6 32.8 28

Abyssinian mustard Brassica carinata - - 3.1 - 1.0 9.7 16.8 16.6 0.7 - 42.5 29 13

Soybean Glycine max - - 10.1 - 4.3 22.3 53.7 8.1 - - - 30

Soybean Glycine max

GMOa,b

- - 3.5 0.1 2.8 22.7 60.3 9.8 0.2 0.2 - 31

Soybean Glycine max

GMOa,c

- 0.1 10.9 0.1 5.7 27.5 51.5 3.0 0.5 0.4 - 31

Soybean Glycine max

GMOa,d

- 0.1 23.8 0.7 3.8 15.4 44.1 11.0 0.4 0.6 - 31

Soybean Glycine max

GMOa,e - - 8.0 0.1 24.7 17.2 39.2 8.3 1.5 0.7 - 31 Palm Elaeis guineensis 0.3 1.2 44.3 - 4.3 39.3 10.0 - - - - 32 Palm Elaeis oleifera - 0.2 18.7 1.6 0.9 56.1 21.1 - - - - 32

Palm kernel Elaeis

guineensis 50.1 15.4 7.3 - 1.8 14.5 2.4 - - - - 32

Palm kernel Elaeis

oleifera 29.3 25.7 10.1 - 1.8 26.4 4.5 - - - - 32

Palm kernel Aiphanes

acanthophylla 41.5 20.5 10.2 - 3.4 15.8 7.4 - - - - 33

Palm kernel Buttia capitata

39.2 6.4 4.2 - 3.0 11.9 3.5 - - - - 33

Palm oleinf 0.3 1.2 40.6 0.2 4.3 41.9 11.9 0.4 0.4 - - 34 Palm stearinf 0.3 1.5 61.1 0.1 4.8 25.8 6.5 0.4 0.5 - - 34 Sunflower Helianthus annuus - - 5.2 0.1 3.7 33.7 56.5 - - - - 30 Sunflower Helianthus annuus GMOa,g - - 3.1 0.1 1.5 91.5 2.1 - 0.2 0.7 0.1 31 Sunflower Helianthus annuus GMOa,g - - 4.4 - 4.2 78.3 10.9 - 0.3 1.0 - 35 Sunflower Helianthus annuus GMOa,c - 0.1 7.5 0.1 1.9 13.3 76.0 0.1 0.1 0.4 - 31 Sufflower Carthamus tinctorius - 0.1 6.4 - 2.3 11.6 79.3 - 0.3 - - 36 Groundnut Arachis hypogea - - 11.2 - 3.6 41.1 35.5 0.1 - - - 30

Corn Zea mays - - 11.6 - 2.5 38.7 44.7 1.4 - - - 30

Olive Olea europaea - - 13.8 1.4 2.8 71.6 9.0 1.0 - - - 30 Cottonseed Gossypium hirsutum - - 23.0 - 2.3 15.6 55.6 0.3 - - - 30 15

Linseed Linum usitatissimum - - 5.6 - 3.2 17.7 15.7 57.8 - - - 30 Coconut Cocos nucifera 50.9 21.1 9.5 - 4.9 8.4 0.6 - - - - 37 Sesame Sesamum indicum - - 9.6 0.2 6.7 41.1 41.2 0.7 - - - 30

Rice bran Oryza sativa - - 22.1 - 2.0 38.9 29.4 0.9 - - - 38

Jatropha Jatropha curcas - - 18.5 - 2.3 49.0 29.7 - - - - 39 Karanjaf Pongamia glabra - - 5.8 - 5.7 57.9 10.1 - 3.5 - - 40 Karanja Pongamia pinnata - - 11.7 - 7.5 51.6 16.5 2.7 - - - 41 Neemf Azadirachta indica - - 17.8 - 16.5 51.2 11.7 - 2.4 - - 40 Salf Shorea robusta - - 6.2 - 43.0 41.3 2.1 - 5.5 - - 40

a_{GMO = genetically modified oil;} b_{low saturate;} c_{high linoleic;} d_{high palmitic;} e_{high stearic;} f_{average value;} g_{high oleic}

17

Table 2.5 Iodine value and saponification value of vegetable oils. Oil Saponification value Iodine value Low erucic rapeseed, crude 179.0 109.9

Soybean, crude 190.7 134.6

Palm, crude 200.0 56.9

Palm kernel, crude 246.4 20.7

Sunflower, winterized 190.6 135.4

Sufflower, linoleic-rich, crude 190.3 143.6 Sufflower, oleic-rich, crude 189.3 93.2 Cottonseed, crude 195.2 105.0

Linseed, crude 189.6 188.0

Corn, soap stock 195.9 105.3 Rice bran, crude 180.1 103.9

Coconut, crude 256.4 9.9

Olive, refined 192.0 84.9

Sesame, crude 188.0 109.2

The oil content in soybean seed ranges from 15 to 22% depending on cultivar and environmental conditions during seed maturity. The major fatty acids are oleic (C18:1) and linoleic (C18:2) as can be seen in Table 2.4.

18

2.3.2. Rapeseed Oil, Mustard Oil, and Canola Oil

The word “rape” originates from Latin word “rapum”, which means turnip and belongs to the family Brassica which includes turnip, mustard, cabbage, rutabaga, broccoli, and kale [46]. Rapeseed was among the first domesticated crops and was used as a source of cooking and illumination oil as early as 2000-1500 BC [47]. Brassica crops are among the few vegetable oil sources that can be cultivated in cool climates. The economically important crops in Brassica and Sinapis species include Sinapis alba (white mustard), Brassica nigra (black mustard),

Brassica carinata (Abyssinian mustard), Brassica juncea (brown, oriental, and leaf mustard), Brassica oleracea (cabbage, kale, cauliflower, broccoli), Brassica rapa (turnip, rape), and Brassica napus (rape, rutabaga) [17]. Oilseeds of rapeseed have an oil content of over 40% while

those of mustard have oil content as low as 20% such as that of Sinapis alba [48]. Oil extracted from these seeds majorly contains fatty acids of oleic acid (C18:1), linoleic acid (C18:2), and erucic acid (C22:1). When rapeseed has a erucic acid content higher than 2%, it is called high erucic acid rapeseed (HEAR), while rapeseed having erucic acid content less than 2% is referred to as low erucic acid rapeseed (LEAR) [49]. Erucic acid contained in rapeseed should be avoided in daily diets. It has been reported that cardiac fat infiltration occurs in experimental animals fed erucic acid and it was concluded that erucic acid is potentially toxic. This compound if fed in large quantities might result in heart lesions [46]. In spite of the adverse nutritional effects of erucic acid in model systems, the effects of erucic acid on humans have not been demonstrated. Nevertheless, the use of rapeseed oil containing a high erucic acid level as edible oil has continuously been objected by many organizations throughout history. The Canadian regulations state that for cooking oil, margarine, salad oil, simulated dairy product, shortening, or food

19

resembling margarine or shortening, the erucic and cetoleic acid may not exceed 5% of the total fatty acid [50].

Erucic acid biosynthesis is through the elongation of oleic acid. In brief, erucic acid is formed by an addition of a two-carbon fragment to oleic acid to form eicosenoic acid (C20:1), followed by an addition of another two-carbon fragment to eicosenoic acid to form erucic acid [51]. In the case of low erucic acid rapeseed (LEAR) such as Brassica napus (Canola Oil or Canadian Brassica), the gene that codes for the fatty acid elongation enzyme is missing leading to the accumulation of the precursor fatty acid, i.e., oleic acid. The level of erucic acid can be selected to range from less than 1% to over 60%. The percentage of erucic acid in Canadian LEAR declined between 1980 and 1989 [52]. The advent of low erucic acid rapeseed (LEAR) lead to the accepted use of LEAR for food; however, HEAR could be used in other industries such as fuel, lubricating oil, oleochemicals, and biopolymer production. In 1974, the so-called “double-low” rapeseed, that is rapeseed low in erucic acid and glucosinolate content, has become commercially available in Canada. The Canola Council of Canada trademarked the name “canola” for LEAR since this “double low” rapeseed became the major vegetable oil used in the Canadian diet. Canola oil is the fully refined, bleached, and deodorized edible oil obtained from

Brassica napus or Brassica rapa with low levels of both erucic acid and glucosinolate content.

Under the USA code of Federal Regulation Title 21, the erucic acid content in canola oil shall not exceed 2% of the component fatty acids [49].

20

2.3.3. Palm Oil

Oil palm is believed to have originated in Africa, but is cultivated most intensively in Southeast Asia especially Malaysia and Indonesia that together account for around 80% of the total world production. Palm first received its botanical name from Jacquin in 1763 as Elaeis

guineensis [53]. The word Elaeis is derived from the Greek word elaion, meaning oil, while guineensis implies its origin in the Guinea coast. The genus Elaeis includes Elaeis guineensis

originating in Africa, Elaeis oleifera originating in Central and South America, and Elaeis odora, previously known as Bercella odora, which is not cultivated. Elaeis guineensis is currently the main commercially grown species in Malaysia because it gives the highest yield per bunch while oil from Elaeis oleifera is more unsaturated and yields less oil. The fruit contains shell and one, two, or three kernels. The seed consists of layers of oily endosperm surrounded by a brown testa covered with a network of fibres. Palm affords the highest oil production per area per year (Table 2.1). There are generally two types of oil derived from palm, including palm oil derived from the mesocarp and palm kernel oil from the kernel inside the testa [22,54]. Palm oil is more saturated than soybean oil and rapeseed oil because its major fatty acids include palmitic (C16:0), stearic (C18:0), oleic (C18:1), and linoleic (C18:2) as shown in Table 2.4. Palm kernel oil is more saturated than palm oil as it mainly contains lauric (C12:0), myristic (C14:0), and oleic (C18:1) acids. Palm oil can be fractionated at ambient temperature (25-30°C) into palm olein or oleic-rich oil (liquid fraction) and palm stearin or stearic-oleic-rich oil (solid fraction). Due to the saturated fatty acids contained in this oil, it has superior oxidative stability compared to other vegetable oils.

21

2.3.4. Sunflower Oil

The genus Helianthus annuus is the botanical name for sunflower, a member of Compositae, or flowering plants, grown throughout the world. The genus name stems from the Greek words helios, meaning sun, and anthos, meaning flower. Sunflower originated in Southwest United States and Mexico [45]. Sunflowers are cultivated both for ornamental and consumption purposes. Sunflower seeds are edible and often crushed to extract oil. The major fatty acids in sunflower oil are oleic (C18:1) and linoleic (C18:2). Sunflower is considered as one of the most ancient oilseed species as its cultivation can be traced back to 3000 B.C. Prior to the advent of the soybean boom after World War II, sunflower was the major source of vegetable oil.

2.3.5. Rice Bran Oil

Rice bran is obtained when brown rice is pearled to produce white rice. The bran and yarn removed by pearling are the main source of rice oil. Lipid droplets can be extracted from rich bran using an extruder, expander, and expeller to form a bran flake or pellet followed by solvent (usually hexane) extraction in an extraction bed. The majority of oil components is triacylglyceride (TAG) with palmitic (C16:0), oleic (C18:1), and linoleic (C18:2) acids as the major fatty acids. Diacylglyceride (DAG), monoacylglyceride (MAG), and sterols may be present in minor amounts. Rice bran oil is used widely in Asian countries due to its delicate flavour and odour. It is recently gaining interest as healthy oil since it reduces serum cholesterol [55].

22

2.3.6. Jatropha Oil

Jatropha curcus is a member of the Euphorbiaceae family. It originated in America, but

is cultivated mainly in Asia, especially India. Jatropha is well adapted to both arid and semi-arid conditions and sheds its leaves in order to survive during drought seasons [56]. It can be grown on non-cultivated degraded wasteland and is considered one of the most promising feedstock materials for biodiesel production [57]. Although Jatropha plants have minimal nutritional requirements, cultivation of Jatropha under acidic soil requires additional nutrients such as calcium and magnesium due to its preference for alkaline soils. Oil derived from Jatropha is non-edible due to curcin, a toxic compound, found in the seeds. Its oil content ranges from 35-40% in seed and 50-60% in kernel with oleic (C18:1) and linoleic (C18:2) as its major fatty acids.

2.3.7. Karanja Oil

Karanja Pongamia pinnata is a member of Leguminaceae family. It is an oil seed bearing tree, native to humid and subtropical environments such as those encountered in Philippines, Indonesia, Malaysia, Myanmar, Australia, India, and United States. It is highly tolerant to salinity and can be cultivated on degraded wasteland and a variety of soil types ranging from clay to sandy or stony. In addition, it plays important role in improving soil quality so that the land exhausted of nutrients can be reused for agricultural purposes [56]. The oil droplet extracted from Karanja appears yellowish orange to brown and is not edible due to the presence of toxic flavonoids [58]. Its oil content varies from 9-46% with oleic (C18:1) and linoleic (C18:2) being the major fatty acids.

23

2.3.8. Used Cooking Oil

The properties of used cooking oil depend highly on the origin and history of the oil. The origin of used cooking oil determines its fatty acid compositions. The history or duration that the oil exposed to water, heat, food, micro-organisms and oxygen during cooking determines its physical and chemical properties such as viscosity, water content, free fatty acid content, and the presence of polymerized and oxidized compounds.

Oil degradation during cooking occurs through three main reactions: thermolytic, oxidative, and hydrolytic reactions. Thermolytic reactions occur in the absence of oxygen and saturated fatty acids, forming alkanes, fatty acids, ketones, esters, and diacylglycerides that are decomposed at high temperatures. In addition, dimeric compounds are major products of thermolytic reactions of unsaturated fatty acids. Dimerization and polymerization of unsaturated fatty acids take place via Diels-Alder reactions. For example, a reaction between conjugated diene from linoleate and oleate can take place to produce a tetra-substituted cyclohexene [59]. In the presence of oxygen, oxidative and nonoxidative reactions will occur simultaneously.

Oxidative reactions occur in a series of initiation, propagation, and termination steps as shown in Figure 2.1. The initial step involves abstraction of hydrogen from unsaturated fatty acid to form a free radical (R·) followed by a reaction of the radical with molecular oxygen to form peroxide radicals (ROO·). The propagation phase involves intermolecular interactions, whereby the peroxide radical abstracts hydrogen from an adjacent molecule, which gives rise to hydroperoxides (ROOH) and a new free radical.

24 Initiation: + •

₊

→

R

H

RH

₊

_O

_→

_ROO

R

ROOH

H

ROO

+

→

₊

→

RO

OH

ROOH

ROH

H

RO

+

→

OH

+

H

→

H

O

25

Carbon-hydrogen bond dissociation energies of fatty acid are lowest at bisallylic, followed by allylic positions (see Figure 2.2). It is reported that lower bond energies for bisallylic and allylic hydrogens are 75 and 88 kcal/mol, respectively, while those of methylene hydrogens are 100 kcal/mol [60]. As a result, hydrogens at bisallylic and allylic locations are favoured sites for proton abstraction by peroxide radicals. Once formed, hydroperoxides tend to proceed toward further oxidation degradation, leading to secondary oxidation derivatives such as aldehydes, acids, and other oxygenates [59]. Hydrolytic reactions take place between the oil and water formed during food preparation. Formations of DAG, MAG, FFA, and glycerol are main derivatives from hydrolysis of TAG [61].

R-CH2- CH2-CH=CH-CH2-CH =CH-R׳

Figure 2.2 Carbon-hydrogen bond positions in fatty acids. alkyl _bisallylic

26

As a result of a combination of these reactions during food preparation, various reaction derivatives are formed leading to an increase in polar content of the oil. It is advised that used cooking oil no longer be used for edible purposes when its polar content exceeds 25% [62]. Therefore, used cooking oils were sold commercially as animal feed. However, in 2002 the European Union (EU) enforced a ban on these waste oils as animal feed because various harmful compounds are formed in used cooking oil during food preparation. When used cooking oil is mixed in feeding meals for domestic animals, these harmful compounds could be returned into the food chain through animal meats [63]. This concern has further raised interest in utilizing used cooking oil as feedstock for biodiesel production.

An obvious advantage of used cooking oil over other vegetable oils is its cheaper price. The prices of soybean, sunflower, yellow grease (FFA <20 wt.%), and brown grease (FFA >20 wt.%) are 18, 20, 9, and 5 to -5 cents/lb, respectively [64]. The negative value of brown grease price implies the cost associated with waste treatment prior to dumping. The availability of used cooking oil as a feedstock for biodiesel production is highly related to area population. Yellow grease generated in Canada is roughly equivalent to 4 kg production per person per year, meaning that approximately 124 Kilotonne of yellow grease is produced annually in Canada [65]. In comparison to used cooking oil, biodiesel produced from fresh vegetable oils would be pricier. Zhang et al. [66] reported that on average a $0.01/kg increase in canola seed cost would result in $0.03/kg increment in biodiesel prices and that raw material cost is responsible for approximately 70-95% of biodiesel production cost when fresh vegetable oil is used as feedstock. Therefore, the use of used cooking oil for feedstock for biodiesel production attracts many biodiesel producers due to its economical benefits.

27 2.4. Biodiesel Production

Transesterification is the most common method used to reduce viscosity of vegetable oils and produce biodiesel [7]. In addition to transesterification of TAG, biodiesel (FAAE) can be produced from free fatty acid (FFA) through esterification. Since ester is characterized by the RCOOR group (R = alkyl group), TAG is a type of ester and the reaction that converts TAG into biodiesel is known as transesterification (transforming ester). In contrast, FFA is not an ester and therefore the reaction to produce biodiesel from FFA is called esterification (making ester). Transesterification is the reaction between glycerides with short chain alcohols and is comprised of three consecutive reactions starting from TAG to DAG to MAG to glycerol, respectively (see Figure 2.3). In each step, the reaction consumes one mole of alcohol and produces one mole of ester. In total one mole of TAG reacts with three moles of alcohols to produce three moles of ester (biodiesel) and one mole of glycerol. In general, the reaction performance is influenced by various parameters such as type of alcohol, alcohol to oil molar ratio, FFA and water content, reaction temperature, reaction duration, and catalyst type. These parameters will be discussed in the following sections.

2.4.1. Effects of Free Fatty Acid and Water Content

Free fatty acid and water content in the starting materials can significantly affect ester yield and glyceride conversion in alkali-catalyzed transesterification. All starting materials including lipid feedstock, alcohol, and catalyst should be substantially anhydrous. Prolonged contact with atmospheric air of alkali catalysts will reduce catalyst efficacy through the catalyst’s interaction with moisture and carbon dioxide in air. Also, it is critical that feedstock used in alkali-catalyzed

28

transesterification should contain free fatty acid (FFA) less than 0.5 wt.% [7]. The higher the acidity of oil, the lower is the conversion and yield in transesterification. If FFA is contained in the starting oil, extra alkali catalyst is needed to neutralize the FFA. The reaction between alkali catalyst and FFA would result in catalyst consumption as well as soap formation and water and is referred to saponification (see Figure 2.4a). Another example of saponification during transesterification is when water is present, it favours hydrolysis of glycerides to form soap and glycerol (see Figure 2.4b). In addition, water can promote hydrolysis of ester to form FFA, which lowers ester yield (see Figure 2.4c). Soap formed during saponification causes increased viscosity or gel formation, which interferes with the transesterification reaction as well as glycerol separation [58]. Ma et al. [67] studied the effects of FFA and water on transesterification of beef tallow using sodium hydroxide and sodium methoxide as a catalyst. It was reported that when 0.6% FFA was added, the yield of beef tallow methyl ester is minimal. Additional water present in the reaction mixture intensely diminished the ester yield. They concluded that FFA and water content should be maintained below 0.5 and 0.06 wt.%, respectively. Low quality feedstocks such as used cooking oil are attractive due to cheaper price. However, these feedstocks usually contain high amounts of FFA and water due to prolonged exposure of heat and contaminated moisture from food. Therefore, direct alkali-catalyzed transesterification of these oils is not applicable. Pre-treatment of these oils to remove FFA and water is usually required. Alkali refining is usually used in oil processing in order to remove FFA from oils [68]. In this process, 12% aqueous sodium hydroxide solution is required to neutralize FFA and to precipitate phosphatides. The treatment temperature and duration can be either 90°C for few seconds (short-mix process) or 40°C for 15 minutes (long-mix process).

29

Figure 2.3 Scheme for step-wise transesterification reaction. DIGLYCERIDE  H2C  HC H2C  OH O  O  C C  R2 R3  O₌  O₌  +  CH3OH  +

TRIGLYCERIDE  METHANOL  METHYL ESTER 

H2C    HC  H2C  O  O  O  C  C  C  R1  R2  R3  O ₌  O ₌  O ₌  C  CH3O  R1  O ₌  MONOGLYCERIDE  +  CH3OH  +

DIGLYCERIDE  METHANOL  METHYL ESTER 

C  CH3O  R2  O ₌  H2C    HC  H2C  OH  O  O  C  C  R2  R3  O ₌  O ₌  H2C  HC H2C  OH OH  O  C  R3  O₌  GLYCEROL  +  CH3OH  +

MONOGLYCERIDE  METHANOL  METHYL ESTER 

C  CH3O  R3  O ₌  H2C    HC  H2C  OH  OH  O  C  R3  O ₌  H2C  HC H2C  OH OH  OH 

30

Figure 2.4 Hydrolysis and saponification during transesterification: a) saponification of free fatty acid; b) saponification of triacylglyceride; c) hydrolysis of methyl ester.

Reaction A:  Reaction B:  Reaction C:  C  KO  R  O = +  KOH +

FREE FATTY ACID  BASE  WATER  SOAP 

H2O C  HO  R  O ₌  C  3 KO  R  O = +  3 KOH H2C  HC H2C  OH  OH  OH  + 

TRIGLYCERIDE  BASE  GLYCEROL  SOAP 

H2C    HC  H2C  O  O  O  C  C  C  R  R  R  O = O = O = C  CH3O  R  O ₌  C  HO  R  O ₌  +  H2O  +

METHYL ESTER  WATER METHANOL FREE FATTY ACID 

31

The oil-soap mixture is then centrifuged to separate the aqueous phase containing water, soap, and precipitated phosphatides. The treated oil usually has FFA reduced to <0.05% and phosphorus to <2 ppm. The disadvantage of this process is the generation of waste water.

FFA can also be removed from vegetable oils through distillation [69]. The distillation process should be performed under vacuum conditions in order to lower the operating temperature. If the operating temperature is too high, glycerides will degrade to generate more acids. The distillation temperature ranged from 100-180°C. However, this approach is less preferred due to the additional cost associated with the distillation step. Alternatively, a two-step acid-alkali esterification-transesterification process can be used [70]. In the first step, FFA is esterified with a short-chain alcohol with acid catalyst to produced ester. Since FFA is converted into ester in the first step, an alkali catalyst can be used in transesterification in the second step. A solid acid catalyst was also reported for simultaneous catalysis of esterification of FFA and transesterification of glycerides [71]. However, further research and development is required to improve conversion and ester yield.

2.4.2. Effects of Alcohol

Stoichiometrically, one mole of TAG requires three moles of alcohol in transesterification. However, due to the reversible nature of the reaction, excess alcohol is usually used in transesterification in order to shift the reaction to the product side. In general, 98% conversion can be achieved at 6:1 alcohol to oil ratio for an alkali-catalyzed reaction and an increase in alcohol used in the reaction does not increase conversion any further [72]. However, an optimum alcohol to oil ratio can be different depending on oil quality and the type of

32

vegetable oil used. It was reported that a maximum of 92% conversion was achieved using 10:1 methanol to oil ratio for biodiesel preparation from Karanja oil [73]. Leung and Guo [74]

reported that 98% ester content can be obtained from transesterification of canola oil using 6:1 alcohol to oil ratio while transesterification of used cooking oil requires 7:1 alcohol to oil ratio to obtain 94% ester content. Transesterification of Cynara cardunculus L. oil requires 12:1 ethanol to oil ratio as an optimum ratio while an increase in ethanol to oil ratio to 15:1 decreases ester content [75]. Rashid and Anwar [76] also reported that a further increase in alcohol used in transesterification of rapeseed oil beyond its optimum ratio (6:1 in this case) would result in reduced ester yield. When too much alcohol is used in transesterification, the polarlity of the reaction mixture is increased, thus increasing solubility of glycerol back into the ester phase and promoting the reverse reaction between glycerol and ester or glyceride, thereby, reducing ester yield. Acid catalyzed reaction requires a higher alcohol to oil molar ratio (30:1), compared to alkali-catalyzed reactions [77-79]. In some cases, the alcohol to oil ratio is increased to 245:1 to obtain 99% conversion [80].

The type of alcohol used in transesterification can also affect reaction performance. Methanol is most commonly used in transesterification, mainly because of its economical benefit [7]. The disadvantages of using methanol are dependency on petroleum sources and a low solubility of TAG in methanol. To illustrate the immiscible behaviour of TAG in methanol, it is reported that a minimum mixing time of 3 minutes is required to sustain methanolysis of soybean oil [81]. A lag time of 2-3 minutes during methanolysis of soybean oil and sunflower oil is also reported [77,82]. This immiscibility behaviour is often referred to as mass transfer resistance or mass transfer limitation which can be overcome by several methods including the use of rigorous